Jutta Hämmerle-Uhl

Cancelable Iris Biometrics Using Block Re-mapping and Image Warping

The concept of cancelable biometrics provides a way to protect biometric templates. A possible technique to achieve such protection for iris images is to apply a repeatable, non-reversible transformation in the image domain prior to feature extraction. We applied two classical transformations, block re-mapping and texture warping , to iris textures obtained from the CASIA V3 Iris database and collected experimental results on the matching performance and key sensitivity of a popular iris recognition method.

Watermarking as a means to enhance biometric systems: a critical survey

Watermarking is discussed as possible means to enhance biometric systems. Application scenarios for the employment of watermarks as found in literature are discussed and analysed with respect to required watermark properties, possible attacks, and eventual (cryptographic) alternatives.

Transforming rectangular and polar iris images to enable cancelable biometrics

To enable cancelable biometrics, we apply two classical transformations, block re-mapping and texture warping, in two variants to iris image data: first, the transformations are applied to rectangular iris imagery prior to iris detection and iris texture unwrapping, and second, the transformations are applied to polar iris images after the generation of the corresponding iris texture patch. The CASIA V3 Iris database is used and experimental results on the matching performance and key sensitivity of a popular iris recognition method employing the cancelable transforms are given.

Experimental study on the impact of robust watermarking on iris recognition accuracy

Watermarking has been suggested as a means to improve security of biometric systems or to add additional functionalities to such systems. We experimentally investigate the impact of applying a set of blind robust watermarking schemes on the recognition performance of two iris recognition algorithms. We find that different watermarking schemes result in a very different amount of impact rendering the choice of a particular watermarking scheme an important issue to be considered in the investigated context.

Attack against robust watermarking-based multimodal biometric recognition systems

Several multimodal biometric schemes have been suggested in literature which employ robust watermarking in order to embed biometric template data into biometric sample data. In case robust embedding is used as the sole means of security, tampering attacks can be mounted. The results of a corresponding attack against a multimodal iris recognition scheme show, that in this environment either semi-fragile watermarking or additional classical cryptographic means need to be applied to secure the system against the demonstrated attack.

Improving Compressed Iris Recognition Accuracy Using JPEG2000 RoI Coding

The impact of using JPEG2000 region of interest coding on the matching accuracy of iris recognition systems is investigated. In particular, we compare the matching scores as obtained by a concrete recognition system when using JPEG2000 compression of rectilinear iris images with and without region of interest coding enabled. The region of interest is restricted to the iris texture area plus the pupil region. It turns out that average matching scores can be improved and that the number of false negative matches is significantly decreased using region of interest coding as compared to plain JPEG2000 compression.

Robust watermarking in iris recognition: application scenarios and impact on recognition performance

Watermarking has been suggested as a means to improve security of biometric systems. We discuss application scenarios for resolving various security issues in this context. As embedding watermarks changes biometric data, degraded recognition performance using such data may be expected. We experimentally investigate the impact of applying a set of blind robust watermarking schemes on the recognition performance of two iris recognition algorithms. We find that different watermarking schemes result in a very different amount of impact rendering the choice of a particular watermarking scheme an important issue to be considered in the investigated context.

Towards standardised fingerprint matching robustness assessment: the StirMark toolkit -- cross-database comparisons with minutiae-based matching

We propose to establish a standardised tool in fingerprint recognition robustness assessment, which is able to simulate a wide class of acquisition conditions, applicable to any given dataset and also of potential interest in forensic analysis. As an example, StirMark image manipulations (as being developed in the context of watermarking robustness assessment) are applied to fingerprint data to generate test data for robustness evaluations, thereby interpreting certain image manipulations as being highly related to realistic fingerprint acquisition conditions. Experimental results involving three different minutiae-based fingerprint matching schemes applied to FVC2004 data underline the need for standardised testing and a corresponding simulation toolset.

Endoscope distortion correction does not (easily) improve mucosa-based classification of celiac disease.

Distortion correction is applied to endoscopic duodenal imagery to improve automated classification of celiac disease affected mucosa patches. In a set of six edge- and shape-related feature extraction techniques, only a single one is able to consistently benefit from distortion correction, while for others, even a decrease of classification accuracy is observed. Different types of distortion correction do not lead to significantly different behaviour in the observed application scenario.

Improved endoscope distortion correction does not necessarily enhance mucosa-classification based medical decision support systems

Distortion correction in two variants is applied to endoscopic duodenal imagery aiming at an improvement of automated classification of celiac disease affected mucosa patches. In a set of heterogeneous feature extraction techniques, only geometry and shape related ones are able to benefit from distortion correction, while for others, even a decrease of classification accuracy is observed. Different types of distortion correction do not lead to significantly different behaviour in the observed application scenario.